Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Unveiling Unicode's Unseen Underpinnings in Undermining Authorship Attribution

Robert Dilworth

TL;DR

This paper investigates how Unicode zero-width steganography can augment adversarial stylometry to obscure authorship attribution, balancing privacy with potential misuse. It presents a multi-technique framework—combining imitation, translation, obfuscation, and zero-width steganography—embodied in the TraceTarnish pipeline, and provides a Python-based proof of principle for encoding hidden signals in text. The authors define evaluation metrics (soundness, safety, sensibility) and report preliminary results using Burrows' Delta to gauge attribution disruption, while discussing ethical, legal, and real-world implications. The work highlights the potential for privacy-enhancing mechanisms in digital communication and outlines concrete directions for future experiments, corpus design, and code dissemination.

Abstract

When using a public communication channel -- whether formal or informal, such as commenting or posting on social media -- end users have no expectation of privacy: they compose a message and broadcast it for the world to see. Even if an end user takes utmost precautions to anonymize their online presence -- using an alias or pseudonym; masking their IP address; spoofing their geolocation; concealing their operating system and user agent; deploying encryption; registering with a disposable phone number or email; disabling non-essential settings; revoking permissions; and blocking cookies and fingerprinting -- one obvious element still lingers: the message itself. Assuming they avoid lapses in judgment or accidental self-exposure, there should be little evidence to validate their actual identity, right? Wrong. The content of their message -- necessarily open for public consumption -- exposes an attack vector: stylometric analysis, or author profiling. In this paper, we dissect the technique of stylometry, discuss an antithetical counter-strategy in adversarial stylometry, and devise enhancements through Unicode steganography.

Unveiling Unicode's Unseen Underpinnings in Undermining Authorship Attribution

TL;DR

This paper investigates how Unicode zero-width steganography can augment adversarial stylometry to obscure authorship attribution, balancing privacy with potential misuse. It presents a multi-technique framework—combining imitation, translation, obfuscation, and zero-width steganography—embodied in the TraceTarnish pipeline, and provides a Python-based proof of principle for encoding hidden signals in text. The authors define evaluation metrics (soundness, safety, sensibility) and report preliminary results using Burrows' Delta to gauge attribution disruption, while discussing ethical, legal, and real-world implications. The work highlights the potential for privacy-enhancing mechanisms in digital communication and outlines concrete directions for future experiments, corpus design, and code dissemination.

Abstract

When using a public communication channel -- whether formal or informal, such as commenting or posting on social media -- end users have no expectation of privacy: they compose a message and broadcast it for the world to see. Even if an end user takes utmost precautions to anonymize their online presence -- using an alias or pseudonym; masking their IP address; spoofing their geolocation; concealing their operating system and user agent; deploying encryption; registering with a disposable phone number or email; disabling non-essential settings; revoking permissions; and blocking cookies and fingerprinting -- one obvious element still lingers: the message itself. Assuming they avoid lapses in judgment or accidental self-exposure, there should be little evidence to validate their actual identity, right? Wrong. The content of their message -- necessarily open for public consumption -- exposes an attack vector: stylometric analysis, or author profiling. In this paper, we dissect the technique of stylometry, discuss an antithetical counter-strategy in adversarial stylometry, and devise enhancements through Unicode steganography.

Paper Structure

This paper contains 39 sections, 7 figures, 3 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Problem Statement
  3. Paper Structure
  4. The Dilemma of Adversarial Stylometry: Exploring Arguments and Implications
  5. A Case for Stylometry
  6. A Case Against Stylometry: Necessitating the Need for Adversarial Stylometry
  7. Closing Remarks: To "Stylometry," or Not to "Stylometry," That is the Question
  8. Overview of Key Concepts
  9. How Zero-Width Characters Can Aid Adversarial Stylometry
  10. Embedding Noise or Decoys
  11. Signal Encoding and Feature Diversion
  12. Creating Multiple "Layers" of Style
  13. Potential Benefits, Limitations, and Challenges
  14. Real-World Considerations
  15. Discussion
  16. ...and 24 more sections

Figures (7)

  • Figure 1: Overview of Adversarial Stylometric Methods (Imitation, Translation, Obfuscation, Unicode Steganography?) and Stylometrist Examination
  • Figure 2: Side-by-Side Transformations of "Never consent to, or blindly accept, a privacy policy without first reading it in its entirety" Across Six Conceptual Lenses: Imitation, Translation, Obfuscation, Soundness, Safety, and Sensibility
  • Figure 3: Staining the Canvas: An example of injecting adversarial noise into a unigram using a steganographic payload composed of zero-width Unicode characters
  • Figure 4: TraceTarnish Script: Sample Workflow Visualization
  • Figure 5: TraceTarnish Script: Terminal Output with More Relevant Fast Stylometry Training Data
  • ...and 2 more figures